Meilisearch MCP Server

Instructions

Implementation Reference

• src/tools/vector-tools.ts:94-103 (handler)

  The handler function that performs the core logic of fetching the embedders configuration for the specified index via the Meilisearch API endpoint.

  async ({ indexUid }) => {
    try {
      const response = await apiClient.get(`/indexes/${indexUid}/settings/embedders`);
      return {
        content: [{ type: "text", text: JSON.stringify(response.data, null, 2) }],
      };
    } catch (error) {
      return createErrorResponse(error);
    }
  }

• src/tools/vector-tools.ts:91-93 (schema)

  Input schema defining the required 'indexUid' parameter as a string.

  {
    indexUid: z.string().describe("Unique identifier of the index"),
  },

• src/tools/vector-tools.ts:88-104 (registration)

  Registration of the 'get-embedders' tool using server.tool(), including name, description, input schema, and handler.

  server.tool(
    "get-embedders",
    "Get the embedders configuration for an index",
    {
      indexUid: z.string().describe("Unique identifier of the index"),
    },
    async ({ indexUid }) => {
      try {
        const response = await apiClient.get(`/indexes/${indexUid}/settings/embedders`);
        return {
          content: [{ type: "text", text: JSON.stringify(response.data, null, 2) }],
        };
      } catch (error) {
        return createErrorResponse(error);
      }
    }
  );

• src/index.ts:68-68 (registration)

  Invocation of registerVectorTools on the main MCP server, which registers the 'get-embedders' tool along with other vector tools.

  registerVectorTools(server);